JP6549682B2 - Group recommendation virtual assistant generation - Google Patents

Description

  Some computing devices may provide a user interface. From the user interface, the user can chat, talk or communicate with a virtual computational assistant (e.g., also referred to as "intelligent assistant" or simply "assistant") to provide useful information to the assistant Specific actions can be performed to output, respond to the needs of the user, or assist the user in completing various real-world or virtual tasks. Some assistants may perform actions to simultaneously assist a group of users, such as providing recommendations to the group of users based on the preferences of the group members.

  In general, the techniques of the present disclosure are directed to virtual computing assistants (e.g., also referred to as "intelligent assistants" or simply referred to as "assistants") associated with the users of the group, but to the private information of the individual users. It may be possible to provide recommendations for the group based on the user's private information without indicating to other users of the group. For example, the calculation assistant may also generate a list of original recommendations based on available information for the user (eg, public information for the user or information the user has agreed to make available for recommendation generation) Good. The original recommendation list may be the general interest of the user to the group, and one or more independent users of the group may have private preferences in favor of or against some of the recommendations. May be included. Thus, the original recommendation list may be of more interest to the group if the assistant has access to private information for the user indicating the private preference when generating the recommendation. However, it may not be desirable for an assistant associated with a particular user to access another user's other private information and preferences.

  According to one or more techniques of the present disclosure, each assistant evaluates the recommendations from the original list for inclusion in the pruned recommendations list based on the respective private information for the associated user. You may Each assistant determines that the recommendation will not be included in the cutaway list if the respective private information for the associated user indicates that the associated user has a negative preference for the recommendation ( That is, it may be knocked out. The assistant may then provide (i.e., display or output) the clipped list to the user. In this way, the respective assistants may provide higher quality group recommendations that take into account the independent preferences of each of the users in the group while preserving the privacy of the associated user's private information.

  In one example, the method responds to receiving a request for recommendations to a group of users and based on the first privacy level data for users of the group, the original recommendation list for the group Including the step of generating. In this example, the method further comprises, based on the respective calculation assistants associated with the users of the group and the respective second privacy level data for the users of the group, a cut recommendations list for the group. Evaluating the recommendation from the original recommendation list to include, wherein the second privacy level is more restricted than the first privacy level. In this example, the method further comprises the step of outputting the cut-off recommendation list for display to the user of the group in response to the cut-off recommendation list including at least one recommendation .

  In another example, the computing device, when executed with the at least one processor, responds to the at least one processor receiving a request for a recommendation to the group of users and the user of the group Create an original recommendation list for the group based on the first privacy level data for the group, and output the original recommendation list to the respective calculation assistants associated with the users of the group; Should recommendations from the original recommendation list be included in the cut-off recommendations list for the group based on the respective second privacy level data for the users of the group from each calculation assistant associated with the user? Receive an indication of whether or not the cut-off recommendation list is at least At least one memory comprising instructions for providing an assistant configured to output the clipped recommendations list for display to the users of the group in response to including the recommendations; The second privacy level comprises a memory that is more restrictive than the first privacy level.

  In another example, when executed, one or more processors of the computing device are responsive to receiving a request for a recommendation to a group of users and a first privacy to the users of the group Based on the level data, generate an original recommendation list for the group, and output the original recommendation list to each calculation assistant associated with the user of the group, and each associated with the user of the group Receive an indication from the calculation assistant in step s based on the respective second privacy level data for the users of the group whether or not the recommendations from the original recommendation list should be included in the pruned recommendations list for the group , In response to the cut-off recommendation list including at least one recommendation, A non-transitory computer readable storage medium storing instructions for causing a user of the group to output a dismantled recommendation list, the second privacy level being higher than the first privacy level. A limited, non-transitory computer readable storage medium is provided.

  The details of one or more examples are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the present disclosure will be apparent from the description, the drawings, and the appended claims.

FIG. 1 is a conceptual diagram illustrating an example system that executes an example virtual assistant, in accordance with one or more aspects of the present disclosure. FIG. 16 is a block diagram illustrating an example computing device configured to perform an example virtual assistant, in accordance with one or more aspects of the present disclosure. FIG. 16 is a block diagram illustrating an example computing system configured to execute an example virtual assistant, in accordance with one or more aspects of the present disclosure. 5 is a flowchart illustrating exemplary operations performed by one or more processors executing an exemplary virtual assistant, in accordance with one or more aspects of the present disclosure. 5 is a flowchart illustrating exemplary operations performed by one or more processors executing an exemplary virtual assistant, in accordance with one or more aspects of the present disclosure.

  FIG. 1 is a conceptual diagram illustrating an example system that executes an example virtual assistant, in accordance with one or more aspects of the present disclosure. System 100 of FIG. 1 includes an assistant server system 160 in communication with search server system 180 and computing devices 110A-110N (collectively, "computing devices 110") via network 130. Although system 100 is illustrated as being distributed among assistant server system 160, search server system 180, and computing devices 110, in other examples, features and techniques belonging to system 100 may be computing devices. It may be implemented internally by 110 local components. Similarly, assistant server system 160 may include specific components and may implement various techniques belonging to search server system 180 and / or computing device 110 in the following description.

  Network 130 may represent any public or private communication network, eg, cellular, Wi-Fi, and / or other types of networks for transmitting data between computing systems, servers, and computing devices. . The assistant server system 160 exchanges data with the computing devices of the computing device 110 via the network 130 and accesses the computing device 110 when the computing device 110 is connected to the network 130 It may provide possible virtual assistant services. The assistant server system 160 may exchange data with the search server system 180 via the network 130 to provide the search service provided by the search server system 180. The computing device 110 may exchange data with the search server system 180 via the network 130 to access the search service provided by the search server system 180.

  Network 130 is operatively interconnected to thereby provide information exchange between server systems 160 and 180 and computing device 110, one or more network hubs, network switches, network routers, or the like. Any network device may be included. Computing device 110, assistant server system 160, and search server system 180 may send and receive data across network 130 using any suitable communication technology. Computing device 110, assistant server system 160, and search server system 180 may each be operatively connected to network 130 using respective network links. The link connecting computing device 110, assistant server system 160, and search server system 180 to network 130 may be Ethernet or other type of network connection, such connection being wireless and / or wired It may be a connection.

  The assistant server system 160 and the search server system 180 may be any suitable one such as one or more desktop computers, laptop computers, mainframes, servers, a cloud computing system capable of sending and receiving information to and from a network such as the network 130, etc. Represents a powerful remote computing system. The assistant server system 160 hosts (or at least provides access to) a virtual assistant service. Search server system 180 hosts (or at least provides access to) a search service. In some examples, assistant server system 160 and search server system 180 represent cloud computing systems that provide access to their respective services via the cloud.

  Each of computing devices 110 represents an independent mobile or non-mobile computing device. Examples of computing devices 110 include cell phones, tablet computers, laptop computers, desktop computers, servers, mainframes, set top boxes, televisions, wearable devices (eg, computerized watches, computerized eyes, etc. , Computerized gloves etc), home automation devices or systems (eg intelligent thermostats or home assistant devices), personal digital assistants (PDAs), gaming systems, media players, electronic book readers, mobile television platforms, car navigation and An entertainment system or a virtual assistant running or accessing it via a network like network 130 Configured to receive information Te was, any other type of mobile, non-mobile, wearable, include non-wearable computing device.

  The assistant server system 160 and / or the search server system 180 communicate with the computing device 110 via the network 130 to allow the computing device 110 to access the virtual assistant service provided by the assistant server system 160. The providing and / or computing device 110 may be provided with access to search services provided by the search server system 180. In the process of providing a virtual assistant service, the assistant server system 160 communicates with the search server system 180 via the network 130 to provide search results for providing the user with virtual assistant service information for completing the task. You may get it.

  System 100 may provide virtual assistant services to multiple users. The virtual assistant service may be provided by a combination of local and remote components. In the example of FIG. 1, assistant server system 160 includes remote assistant modules 123A-123N (collectively, "remote assistant modules 123") and user information data stores 125A-125N (collectively, "user information data Store 125 "). Each of the remote assistant modules 123 is part of a respective virtual assistant service provided by the assistant server system 160 to a particular user via the network 130 for each user information data store of the user information data store 125. You may maintain it. For example, remote assistant module 123A may maintain user information data store 125A as part of a virtual assistant service that assistant server system 160 provides user A via network 130, and remote assistant module 123B may , The assistant server system 160 may maintain the user information data store 125B as part of a virtual assistant service provided to the user B via the network 130,..., The remote assistant module 123N the assistant server system 160. The user information data store 125 N may be maintained as part of a virtual assistant service that the user provides to the user N via the network 130.

  Computing device 110A includes a user interface device (UID) 112, a user interface (UI) module 120, a local assistant module 122A, and a user information data store 124A. The local assistant module 122A may maintain the user information data store 124A as part of a virtual assistant service executing on the computing device 110A locally for a particular user, such as user A. Other computing devices of computing device 110 may include similar components to computing device 110A. In some examples, each computing device of computing device 110 may include a respective local assistant module and a respective user information data store associated with a respective user. For example, computing device 110B may include local assistant module 122B (not shown) and user information data store 124B (not shown) associated with user B, with computing device 110N being user B. May include a local assistant module 122N (not shown) and a user information data store 124N (not shown). Local assistant modules 122A-122N may be collectively referred to as "local assistant module 122". User information data stores 124A-124N may be collectively referred to as "user information data store 124".

  Modules 120, 122, 123, and 182 are software, hardware, firmware, or hardware, software that reside and / or execute on one of computing device 110, assistant server system 160, or search server system 180. , And a mix of firmware may be used to perform the described operations. Computing device 110, assistant server system 160, and search server system 180 may execute modules 120, 122, 123, and 182 on multiple processors or multiple devices. Computing device 110, assistant server system 160, and search server system 180 may execute modules 120, 122, 123, and 182 as virtual machines running on underlying hardware. Modules 120, 122, 123, and 182 may be implemented as one or more services of an operating system or computing platform. Modules 120, 122, 123, and 182 may be implemented as one or more executable programs of the application layer of the computing platform.

  The UID 112 of computing device 110A may function as an input and / or output device for computing device 110A. The UID 112 may be implemented using various techniques. For example, the UID 112 may be present as a resistive touch screen, a surface acoustic wave touch screen, a capacitive touch screen, a projected capacitive touch screen, a pressure sensitive screen, an acoustic pulse recognition touch screen, or another presence sensitive display technology A sensing input screen may be used to act as an input device.

  The UID 112 may function as an input device using microphone technology, infrared sensor technology, or other input device technology for use in receiving user input. For example, UID 112 may detect voice input that UI module 120 and / or local assistant module 122 processes to complete a task using built-in microphone technology. As another example, UID 112 may include a presence sensitive display that may receive haptic input from a user of computing device 110A. The UID 112 detects an indication of tactile input by detecting one or more gestures from the user (eg, the user touching or pointing to one or more positions of the PSD 112 using a finger or a stylus pen) May.

  The UID 112 may function as an output (eg, display) device to provide output to the user. The UID 112 may be a liquid crystal display (LCD), a dot matrix display, a light emitting diode (LED) display, an organic light emitting diode (OLED) display, an electronic ink, or similar monochrome or color information capable of outputting visible information to a user of the computing device 110A. Any one or more display devices, such as a color display, may be used as an output device. The UID 112 may function as an output device using speaker technology, haptic feedback technology, or other output device technology for outputting information to the user. The UID 112 may provide a user interface (e.g., user interface 114) associated with the virtual assistant provided by the local assistant module 122 and / or the remote assistant module 123A. The UID 112 is a computing platform, operating system, application, and / or service (eg, email, chat, online service, phone, game, etc.) running on and / or accessible from the computing device 110 User interface associated with the other features of.

  UI module 120 may manage user interaction with UID 112 and other components of computing device 110A, including interacting with assistant server system 160 to provide assistant services via UID 112. UI module 120 may provide UID 112 with a user interface (or other example user interface), such as user interface 114, when the user of computing device 110A references output and / or provides input with UID 112. May be output. The UI module 120 and the UID 112 may allow the user to input one or more instructions (eg, voice input, gesture input, etc.) at different times and when the user and the computing device 110A are in different positions. It may be received from the user when interacting with the user interface. The UI module 120 and the UID 112 interpret the input detected by the UID 112 to interpret information regarding the input detected by the UID 112, the local assistant module 122A running on the computing device 110A and / or one or more May be relayed to other associated platforms, operating systems, applications, and / or services, for example, to cause computing device 110A to perform the function.

  UI module 120 may be one or more associated platforms, operating systems, applications running on computing device 110 A and / or one or more remote computing systems such as server systems 160 and 180. And / or may receive information and instructions from the service. Further, the UI module 120 is associated with the computing device 110A and one or more associations being performed on various output devices (eg, speakers, LED indicators, audio or haptic output devices, etc.) of the computing device 110A. Acting as an intermediary between selected platforms, operating systems, applications, and / or services, the output (e.g., graphics, flash of light, sound, tactile response, etc.) is output at computing device 110A It is also good.

  Search module 182 may be a search query or search module 182 automatically generated by search module 182 (eg, based on context information associated with computing device 110A) from assistant server system 160 or computing device 110. A search for information determined to be relevant to the received search query may be performed (eg, as part of a task being performed by a virtual assistant on behalf of the computing device user of computing device 110) . The search module 182 may generate information related to the search query from various information sources (eg, stored locally or remotely at the search server system 180) (eg, weather or traffic conditions, news, stock prices, etc.) Internet searches may be performed based on search queries to identify, sports scores, user schedules, shipping schedules, real estate prices, etc.). After performing the search, the search module 182 may output the information returned from the search (eg, search results) to the assistant server system 160 or the computing device 110 computing device.

  In the example of FIG. 1, the user interface 114 is a virtual assistant service accessed by the computing device 110A (eg, provided by the local assistant module 122 and / or the remote assistant module 123A of the assistant server system 160). A graphical user interface associated with The user interface 114 may be an audible user interface in some instances. As shown in FIG. 1, user interface 114 may be a user of computing device 110A (ie, “user A”), a user of computing device 110B (ie, “user B”), computing device 110C. The text of the conversation between the user of (i.e., "user C") and the virtual assistant running on or accessible to computing device 110 is shown. The user interface 114 may include virtual assistant information in various forms such as audible sounds, vibrations, text, graphics, content cards, images and the like. UI module 120 may cause UID 112 to output user interface 114 based on data received by data UI module 120 from assistant server system 160 via network 130. The UI module 120 is configured to provide the assistant server system 160 and / or the information for providing the information for displaying the user interface 114 (for example, audio data, text data, image data, etc.) in the user interface 114 with the UID 112. Or may be received as input from assistant server system 160 and / or local assistant module 122 along with instructions from local assistant module 122.

  The local assistant module 122 of the computing device 110 and the remote assistant module 123 of the assistant server system 160 automatically execute their respective assistants for the group of users configured to provide the group of users. Each of the same functions described herein may be performed to The remote assistant module 123 and the user information data store 125 represent an exemplary virtual assistant server side or cloud implementation, and the local assistant module 122 and the user information data store 124 are client side of the exemplary virtual assistant Represents a local or local implementation.

  Modules 122 and 123 may each include respective software agents configured to execute as intelligent personal assistants capable of performing tasks or services to an individual, such as a user of the computing device 110 computing device. . Modules 122 and 123 may be user input (e.g., detected by UID 112), location recognition (e.g., based on context), and / or stored locally (e.g., computing device 110, assistant server system 160) Or other information (eg, weather or traffic conditions, news, stock prices, sports scores, user schedules, transportation schedules, etc.) from various sources of information provided via the search service provided by the search server system 180 These tasks or services may be performed based on the ability to access property prices etc.). Modules 122 and 123 may implement artificial intelligence and / or machine learning techniques to automatically identify and complete one or more tasks on behalf of the user.

  The respective assistants provided by the remote assistant module 123 and the local assistant module 122 automatically create, create or maintain personal records of the information, and the respective personal records as user-specific values, respectively. It may be stored in a semantic form structured as information data store 124 and user information data store 125. Data stores 124 and 125, respectively, allow each assistant executed by remote assistant module 123 and local assistant module 122 to quickly access the personal information (eg, a value specific to the user) to It may be possible to complete world tasks, virtual tasks, or respond to the immediate or future needs of the user.

  The assistant modules 122 and 123 may only maintain personal records associated with the user after first receiving explicit permission from the user to do so. Thus, the user may have complete control over how the assistant collects and uses information about the user. For example, before holding personal records associated with user A of computing device 110A, assistant modules 122A and 123A may cause UI module 120 to provide a user interface via UID 112. The user interface selects a box, clicks a button, speaks an audio input, or is unambiguous for assistant modules 122A and 123A to collect and use the personal information of the user by assistant modules 122A and 123A, Request user A to provide the user interface with specific input that is interpreted as a positive consent.

  The assistant modules 122 and 123 may also encrypt or process the information maintained as personal records to remove the user's actual identity prior to storing the personal information in the data stores 124 and 125. Good. For example, the information may be processed by assistant modules 122 and 123 such that any personally identifiable information may be removed from the user's personal record when stored in data stores 124 and 125. Good.

  The assistant modules 122 and 123 may cause the UI module 120 to provide a user interface. From the user interface, the user may modify or remove information from personal records stored in data stores 124 and 125. For example, the user interface may provide an area where the user can provide input to send commands to the assistant modules 122 and 123 to modify or remove specific personal information. In this manner, each of the users of computing device 110 may have complete control over the information held by the respective assistant modules 122 and 123 at data stores 124 and 125, respectively.

  Assistant modules 122 and 123 may allow the user to control how the information in data stores 124 and 125 is used. For example, assistant modules 122 and 123 may tag, mark, or annotate the information in data stores 124 and 125 at various privacy levels. For example, assistant modules 122 and 123 mark some information in data stores 124 and 125 at a first privacy level, and some information in data stores 124 and 125 a second privacy. -You may mark by level. The second privacy level may be more restrictive than the first privacy level. That is, information marked as having the second privacy level may be considered more private than information marked as having the first privacy level. In some instances, the information marked as having the first privacy level may be public information or information that is publicly available.

  The assistant modules 122 and 123 may utilize the information in the data stores 124 and 125 marked at different privacy levels for different purposes in the course of performing different operations. For example, as the information becomes more private, assistant modules 122 and 123 may become more rigid in protecting the information. For example, assistant modules 122 and 123 may indicate or provide information marked as public to other users and / or virtual assistants associated with other users, and information marked as private to other users And / or may be hidden or protected from virtual assistants associated with other users. In this manner, each of the users of computing device 110 may have complete control over how to use the information in data stores 124 and 125 by their respective assistant modules 122 and 123.

  Each entry in the personal records stored by data stores 124 and 125 can be visited or analyzed quickly by assistant modules 122 and 123 to understand the user's needs and assist the user in completing the task. It may be associated with a pre-defined schema where the needs of the information assistant modules 122 and 123 at the current point in time can be found. When personal information is recorded as one or more values specific to the user, the assistant modules of assistant modules 122 and 123 associated with the user are stored in data stores 124 and 125 associated with the user. Information can be used quickly to complete the task. If there are no tasks in progress, the assistant modules 122 and 123 may give the user (e.g., via the user interface 114) an example of how the assistant can use the information to assist the user in the future. May be provided. The user may later provide input at UID 112 to instruct assistant modules 122 and 123 to forget or modify this information.

  The values stored by the data stores 124 and 125 may be text values (eg, person's name, place's name, other text descriptors of the entity), numbers (eg, age, height, weight, other physiological data) , A pointer to a value specific to the user associated with the data store, or other numeric information associated with the entity (eg, a location in memory for the entity in the user's knowledge graph, a contact in the address book) Location in the memory, etc.). In other words, the user specific value may take many forms and may be specific to the field of personal records defined by the record schema. The value may indicate actual information specific to the user, or may be a reference to a location from which information specific to the user can be retrieved.

  The user of computing device 110 may be in a group. Examples of groups include, but are not limited to, groups relating to social media sites, chat rooms, group chats, ad hoc designation of groups of users, or any other multiple user associations. The user of computing device 110 may participate in the group on its own, may be added to the group (after receiving consent from the user), or may generate the group.

  The assistant modules 122 and 123 may be configured to provide one or more recommendations, also known as group recommendations or multi-user recommendations, to a group of users. For example, the user may request that the assistant modules of assistant modules 122 and 123 associated with the user provide one or more recommendations of the group. Some examples of group recommendations are entertainment (eg, movie recommendations for the group to watch), food (eg, restaurant or bar recommendations for the group to go), travel (eg, travel destination, flight destination, or Recommendations for the location of the tour), and shared products (for example, recommendations for purchasing a car / TV / game console), but not necessarily limited thereto.

  The quality of group recommendations provided by the assistant modules 122 and 123 associated with the requesting user (ie, the user who requested the recommendation) is available to the assistant modules 122 and 123 associated with the requesting user It may relate to the quality of information possible. For example, if the assistant modules 122 and 123 associated with the requesting user use more information about the users in the group, the quality of the recommendation for the result is higher (ie, the recommendation is a member of the group Are more likely to be accepted by Thus, the assistant modules 122 and 123 associated with the requesting user may be able to personalize the independent members of the group in order to generate recommendations likely to be accepted by more members of the group. It may be desirable to use a lot.

  However, the users of the group may not want the information to be accessible by other members of the group or assistant modules 122 and 123 associated with the requesting user. Thus, assistant modules 122 and 123 associated with the requesting user may have limited access to information for users of the group different from the requesting user. Thus, assistant modules 122 and 123 associated with the requesting user may generate the original recommendation list based on the first privacy level data for the users of the group. The first privacy level data may include public information for users of the group, and / or information that users of the group agree to make available for the purpose of group recommendation. Since the original recommendation list is generated based on the restricted data for the users of the group, the original recommendation list is generated by the assistant modules 122 and 123 associated with the requesting user for the other users of the group. It may not be as high quality as having access to private information.

  According to one or more techniques of the present disclosure, the respective assistant modules of assistant modules 122 and 123 associated with the users of the group are associated with a respective second privacy level for the users of the group. Based on the data, the recommendations may be evaluated from the original recommendations list for inclusion in the trimmed recommendations list for the group. For example, assistant modules 122 and 123 (e.g., assistant modules 122A and / or 123A) associated with the first user of the group may be included in the original recommendation list for inclusion in the cut recommendations list for the group. Recommendations from the second user may be evaluated based on second privacy level data for the first user, and assistant modules 122 and 123 associated with the second user of the group (e.g. Modules 122B and / or 123B) evaluate recommendations from the original recommendation list for inclusion in the cut recommendations list for the group based on the second privacy level data for the second user. And so on. The second privacy level data may be more restrictive than the first privacy level data. For example, some data may be marked as second privacy level data by default and / or users of the group may want to mark data as second privacy level data 1 One or more settings may be adjusted.

  The assistant modules of the assistant modules 122 and 123 associated with the users of the group may output the result of the evaluation to the assistant modules 122 and 123 associated with the requesting user. For example, the assistant modules of the assistant modules 122 and 123 associated with the users of the group, for each recommendation of the original recommendation list, put the respective recommendations into the cut recommendations list for the group You may output an indication of whether it should be included. The assistant modules 122 and 123 associated with a particular user may determine that the particular recommendation should not be included in the cut recommendations list. In this case, the second privacy level data for the particular user indicates that the particular user has some negative preference for the recommendation. The data for the user is that the user has a negative preference for the recommendation, the data is such that the user does not accept the recommendation, does not like the recommendation, can accept the recommendation but has another option, or any other negative It may be shown to indicate that it prefers However, the assistant modules 122 and 123 associated with the requesting user may not be provided with justification or an explanation along with the result of the evaluation.

  The assistant modules 122 and 123 associated with the requesting user receive the results of the evaluation and determine whether any recommendations from the original recommendations list should be included in the cut-off recommendations list You may If the assistant module of each of the assistant modules 122 and 123 associated with the user of the group does not determine that the recommendation should not be included in the cut recommendation list, the cut recommendation from the original recommendation list is cut Should be included in the recommended list.

  In response to the cut-off recommendations list including at least one recommendation, the assistant modules 122 and 123 associated with the requesting user select the cut-off recommendations to display to the users of the group. You may output a list. For example, assistant modules 122 and 123 associated with the requesting user may output the cut recommendation list to assistant modules 122 and 123 associated with other users of the group, the assistant module 122 and 123 may be displayed, spoken or output to be provided to the associated user. Since the original recommendation list is not known to the users of the group and the users of the group do not know which user's assistant module did not include the recommendation in the cut-off recommendations list, the users of the group Each assistant module of assistant modules 122 and 123 associated with may effectively have an implicit "deny right" for recommendations from the original recommendation list. In this manner, assistant modules 122 and 123 can provide higher quality recommendations while still preserving the privacy of their associated user's private information.

  In some instances, as in the previous example, the requesting user may be included in the group of users for which the recommendation is generated. For example, the user may request recommendations for the user and one or more other users. Thus, in such an example, the recommendation may be generated based on data for the requesting user. In some instances, the requesting user may not be included in the group of users for which the recommendation is generated. For example, a user may request recommendations for other users (e.g., an administrative assistant may request recommendations for a group of other users). Thus, in such an example, the recommendation may not be generated based on data for the requesting user.

  Some of the advantages provided by the above approach include: 1) utilizing the user's personal information to generate higher quality recommendations for the group that includes the user while maintaining the personal information of the independent user Maintaining privacy, 2) may generate higher quality recommendations to minimize the amount of time, and thus power (eg, battery power), needed to reach group agreement.

  FIG. 2 is a block diagram illustrating an exemplary computing device configured to execute an exemplary virtual assistant in accordance with one or more aspects of the present disclosure. The computing device 210 of FIG. 2 is described below as an example of the computing device of the computing device 110 of FIG. Although FIG. 2 shows only one particular example of computing device 210, many other instances of computing device 210 may be used in other instances and the exemplary computing device A subset of components included in 210 may be included, or additional components not shown in FIG. 2 may be included.

  As shown in the example of FIG. 2, the computing device 210 may be a user interface device (USD) 212, one or more processors 240, one or more communication units 242, one or more input components 244. , One or more output components 246, and one or more storage components 248. The USD 212 includes a display component 202, a presence detection input component 204, a microphone component 206, and a speaker component 208. Storage component 248 of computing device 210 includes UI module 220, assistant module 222, search module 282, one or more application modules 226, and user information data store 224.

  Communication channel 250 may interconnect (physically, communicatively and / or operatively) each of components 212, 240, 242, 244, 246, and 248 for inter-component communication. In some instances, communication channel 250 may include a system bus, a network connection, an inter-process communication data structure, or any other method for communicating data.

  One or more communication units 242 of the computing device 210 transmit and receive network signals over one or more networks (eg, the network 130 of the system 100 of FIG. 1) to generate one or more wired And / or may communicate with external devices (eg, assistant server system 160 and / or search server system 180 of system 100 of FIG. 1) via a wireless network. Examples of communication unit 242 include a network interface card (eg, an Ethernet card), an optical transceiver, a radio frequency transceiver, a GPS receiver, or any other type of device capable of transmitting and receiving information. . Other examples of communication unit 242 may include shortwave radio, cellular data radio, wireless network radio, as well as Universal Serial Bus (USB) controller.

  One or more input components 244 of computing device 210 may receive input. Examples of inputs are tactile, audio and video input. The input component 242 of computing device 210, in one example, detects input from a presence sensitive input device (eg, touch sensitive screen, PSD), mouse, keyboard, voice response system, camera, microphone or human or machine For any other type of device. In some examples, the input component 242 can be one or more sensor components, one or more position sensors (GPS components, Wi-Fi components, cellular components), one or more temperature sensors, one or more Multiple motion sensors (eg, accelerometers, gyroscopes), one or more pressure sensors (eg, barometers), one or more ambient light sensors, and one or more other sensors (eg, infrared proximity) Sensor, a hygrometer, etc.). Other sensors may include heart rate sensors, magnetometers, glucose sensors, hygrometer sensors, olfactory sensors, compass sensors, step counter sensors, to name a few other non-limiting examples.

  One or more output components 246 of computing device 110 may generate output. Examples of outputs are haptic, audio and video output. The output component 246 of the computing device 210 is, in one example, a presence detection display, a sound card, a video graphics adapter card, a speaker, a cathode ray tube (CRT) monitor, a liquid crystal display (LCD), or a human or machine. Including any other type of device for generating output to.

  UID 212 of computing device 210 may be similar to UID 112 of computing device 110 and includes display component 202, presence detection input component 204, microphone component 206, and speaker component 208. The display component 202 may be a screen on which information is displayed by the USD 212, and the presence detection input component 204 may detect an object at and / or near the display component 202. Speaker component 208 may be a speaker where audible information is reproduced by UID 212, and microphone component 206 may detect audible input at and / or near display component 202 and / or speaker component 208.

  Although shown as an internal component of computing device 210, UID 212 may represent an external component that shares a data path with computing device 210 to send and receive input and output. For example, in one example, the UID 212 represents an embedded component of the computing device 210 (e.g., a screen of a mobile phone) disposed within and physically connected to an external packaging of the computing device 210. In another example, the UID 212 is located external to the packaging or enclosure of the computing device 210 and is physically separated from the external components of the computing device 210 (eg, wired and / or wireless data paths) The monitor, projector, etc. shared with the computing device 210).

  As one exemplary range, the presence sensing input component 204 may detect an object such as a finger or stylus that is within 2 inches or less from the display component 202. The presence sensing input component 204 may determine the position (e.g., [x, y] coordinates) of the display component 202 at which the object was detected. In another exemplary range, the presence sensing input component 204 may detect objects 6 inches or less from the display component 202, and other ranges are possible. The presence sensing input component 204 may determine the position of the display component 202 selected by the user's finger using capacitive, inductive, and / or light recognition techniques. In some examples, the presence sensing input component 204 also provides output to the user using haptic, audio, or video stimulation as described for the display component 202. In the example of FIG. 2, PSD 212 may provide a user interface (such as graphical user interface 114 of FIG. 1).

  The speaker component 208 may comprise a speaker built-in to the housing of the computing device 210, and in some instances, to a set of wired or wireless headphones operatively connected to the computing device 210. It may be a speaker built-in. Microphone component 206 may detect an audible input occurring at or near UID 212. Microphone component 206 may implement various denoising techniques to remove background noise and isolate user speech from the detected speech signal.

  UID 212 of computing device 210 may detect two-dimensional and / or three-dimensional gestures as input from a user of computing device 210. For example, the sensor of UID 212 may detect movement of the user within a threshold distance of the sensor of UID 212 (e.g. movement of a hand, an arm, a pen, a stylus, etc.). The UID 212 may determine a two-dimensional or three-dimensional vector representation of the movement, and correlate the vector representation to gesture inputs having multiple dimensions (eg, hand waves, pinches, clapping, pen strokes, etc.). In other words, the UID 212 can detect multi-dimensional gestures without requiring the user to make a gesture on or near the screen or surface on which the UID 212 outputs information for display. Instead, the UID 212 can detect multi-dimensional gestures performed at or near sensors that may or may not be present at or near the screen or surface where the UID 212 outputs information for display. .

  One or more processors 240 may implement functionality and / or execute instructions associated with computing device 210. Examples of processor 240 include an application processor, a display controller, an auxiliary processor, one or more sensor hubs, and any other hardware configured to function as a processor, a processing unit, or a processing device. Modules 220, 222, 226, and 282 may be operable by processor 240 to perform various actions, operations, or functions of computing device 210. For example, processor 240 of computing device 210 may retrieve and execute instructions stored by storage component 248 that cause processor 240 to perform such operational modules 220, 222, 226, and 282. The instructions may cause the computing device 210 to store information in the storage component 248 when executed by the processor 240.

  One or more storage components 248 within the computing device 210 may store information for processing during operation of the computing device 210 (e.g., computing device 210 is a computing device). Data accessed by modules 220, 222, 226, and 282 during execution on device 210 may be stored). In some instances, storage component 248 is a temporary memory, which means that the primary purpose of storage component 248 is not long-term storage. Storage component 248 on computing device 210 may be configured for short-term storage of information as volatile memory, and thus does not hold stored content when powered off. Examples of volatile memory include random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), and other forms of volatile memory known in the art.

  Storage component 248 also includes one or more computer readable storage media in some examples. In some instances, storage component 248 includes one or more non-transitory computer readable storage media. Storage component 248 may be configured to store a larger amount of information than that generally stored by volatile memory. Storage component 248 may further be configured for long term storage of information as non-volatile memory space, and may retain information after power on / off cycles. Examples of non-volatile memory include the form of magnetic hard disks, optical disks, floppy disks, or electrically programmable memory (EPROM) or electrically erasable and programmable (EEPROM) memory. Storage component 248 may store modules 220, 222, 226, and 282 and program instructions and / or information (eg, data) associated with data store 224. Storage component 248 may include modules 220, 222, 226, and 282 and memory configured to store data or other information associated with data store 224.

  UI module 220 may include all the functionality of UI module 120 of computing device 110 of FIG. 1, with computing device 210 facilitating the interaction between the user of computing device 110 and assistant module 222. The same operations as the UI module 120 for managing a user interface (for example, the user interface 114) provided by the USD 212 may be performed. For example, the UI module 220 of the computing device 210 may send information to the assistant module 222 including instructions for outputting an assistant user interface (eg, user interface 114) (eg, displaying or playing sound). You may receive from. The UI module 220 may receive the information from the assistant module 222 over the communication channel 250 and use the data to generate a user interface. UI module 220 may transmit display commands and audible output commands and associated data over communication channel 250 to cause UID 212 to provide a user interface with UID 212.

  In some examples, UI module 220 may receive an indication of one or more user inputs detected at UID 212 and may output information regarding the user inputs to assistant module 222. For example, the UID 212 may detect voice input from the user and send data regarding the voice input to the UI module 220.

  The UI module 220 may send an indication of the voice input to the assistant module 222 for further interpretation. The assistant module 222 may determine based on the speech input that the detected speech input represents a user request for the assistant module 222 to perform one or more tasks.

  The application module 226 may be any of a variety of things executed by and accessible from the computing device 210 that may be accessed by an assistant such as the assistant module 222 providing information to the user and / or performing tasks. Represents independent applications and services. A user of computing device 210 may interact with a user interface associated with one or more application modules 226 to cause computing device 210 to perform a function. There may be numerous examples of application modules 226, such as fitness applications, calendar applications, personal assistants or prediction engines, search applications, map or navigation applications, transport services applications (eg bus or train tracking applications), social media Applications, gaming applications, email applications, chat or messaging applications, Internet browser applications, or any and all other applications that may run on computing device 210.

  The search module 282 of computing device 210 may perform integrated search functionality on behalf of computing device 210. The search module 282 may be invoked by one or more of the UI module 220, the application module 226, and / or the assistant module 222 to perform the search operation instead. When invoked, the search module 282 may perform search functions such as generating search queries and performing searches based on search queries generated across various local and remote information sources. The search module 282 may provide the results of the performed search to the launch component or module. That is, the search module 282 may output search results to the UI module 220, the assistant module 222, and / or the application module 226 in response to the activation command.

  The assistant module 222 may include all features of the local assistant module of the local assistant module 122 of the computing device of the computing device 110 of FIG. The same operation may be performed. In some examples, assistant module 222 may execute locally (eg, at processor 240) to provide assistant functionality. In some examples, assistant module 222 may act as an interface to a remote assistant service accessible to computing device 210. For example, assistant module 222 may be an interface or application programming interface (API) to the remote assistant module of remote assistant module 124 of assistant server system 160 of FIG.

  Similar to the user information data stores 124 and 125 of the system 100 of FIG. , Personal records created by the assistant module 222 may be stored. The assistant module 222 may perform the assistant task depending on the information stored in the data store 224 in addition to any information provided by the search module 282.

  Similar to assistant modules 122 and 123 of FIG. 1, assistant module 222 may be configured to generate recommendations for a group of users. For example, the assistant module 222 may generate an original recommendation list based on first privacy level data for the users of the group, and may generate the original recommendations for evaluation by computational assistants associated with other users of the group. You may output a list of recommendations for Assistant module 222 may receive from the calculation assistant associated with other users of the group an indication of whether recommendations from the original recommendation list should be included in the cut-off recommendation list for the group . In some instances, the indication may be positive. For example, a calculation assistant associated with another user of the group may output to the assistant module 222 a recommendation list from the original recommendation list that has been approved for inclusion in the cropped recommendation list. In some instances, the indication may be negative. For example, a calculation assistant associated with another user of the group may send the assistant module 222 a recommendation list from the original recommendation list that has not been approved for inclusion in the cut recommendation list (ie, another user The recommendation list from the original recommendation list that the calculation assistant associated with is rejected may be output.

  The computational assistant associated with the users of the group may generate the indication based on the respective second privacy level data for the users of the group. As mentioned above, the use of data marked at the second privacy level may be more restricted than data marked at the first privacy level. As mentioned above, the users of the group may be provided with an option to control which data is marked as the second privacy level, and calculations associated with other users of the group The assistant may receive consent to use the data marked at the second privacy level to generate a group recommendation.

  A computational assistant associated with a particular user of the group may determine that the particular recommendation should not be included in the cut-off recommendation list. In this case, the second privacy level data for the particular user indicates that the particular user has some negative preference for the recommendation. The data for the user is that the user has a negative preference for the recommendation, the data is such that the user does not accept the recommendation, does not like the recommendation, can accept the recommendation but has another option, or any other negative It may be shown to indicate that it prefers

  The assistant module 222 may generate the cut recommendation list based on the evaluation performed by the calculation assistant. For example, assistant module 222 may determine whether each recommendation included in the original recommendation list has been rejected by at least one of the calculation assistants. If the recommendations from the original recommendations list were not rejected by any of the calculation assistants, the assistant module 222 may include the recommendations in the cut recommendations list.

  In response to the cropped recommendations list including at least one recommendation (ie, at least one recommendation from the original recommendations list was not rejected by any of the calculation assistants associated with the users of the group If so, the assistant module 222 may output the cut recommendations list for display to the users of the group. As one example, the assistant module 222 may cause the UI module 220 to display, via the display component 202, a GUI that includes the clipped recommendation list. As another example, the assistant module 222 may cause the UI module 220 to output, via the speaker component 208, the synthesized conversation including the cut recommendation list. As another example, the assistant module 222 may output the cut recommendation list to calculation assistants associated with other users of the group. The calculation assistant may then display, speak or output the cut recommendation list to its respective user.

  In some instances, the cut recommendations list may not include recommendations. For example, all recommendations from the original recommendations list may be rejected by at least one calculation assistant of the calculation assistant associated with the user of the group. In some examples, assistant module 222 may inform the user of computing device 210 that assistant module 222 could not generate a recommendation for the group. However, it may be desirable for the assistant module 222 not to generate recommendations for the group, especially if the recommendations are clipped based on the user having some negative preference for the recommendations .

  In accordance with one or more techniques of the present disclosure, the assistant module 222 causes the computational assistant associated with the users of the group to use the weighted second privacy level data for the users of the group. The recommendations from the original recommendations list may be evaluated for inclusion in the improved reaped recommendations list. As an example, a calculation assistant associated with a particular user of the group may have second privacy level data for a particular user of the group such that the particular user has a complete negative preference for the recommendation. In response to determining that it indicates, it may be determined that recommendations from the original recommendation list should not be included in the improved reaped recommendation list. As another example, a calculation assistant associated with a particular user of the group may have second privacy level data for a particular user of the group that the particular user has a partially negative preference for the recommendation. In response to determining that it indicates that it has, it may be determined that recommendations from the original recommendation list should be included in the improved reaped recommendation list.

  The data for the user is that the user has a completely negative preference for the recommendation, the data does not accept the recommendation, the user does not accept the recommendation, never accepts the recommendation, or any other complete May indicate negative preferences. For example, data indicating that the user is a vegetarian may indicate that the user has a completely negative preference if the user is angry about going to the steakhouse. The data for the user is that the user has a partially negative preference for the recommendation and the data allow the user to accept the recommendation but prefer another option, and usually do not like the recommendation but for the group It may be shown to accept recommendations or to indicate any other palliative connotations. For example, data indicating that the user is a vegetarian may indicate that the user has a partially negative preference if it is OK for the user to order a salad.

  The assistant module 222 may generate an improved cropped recommendation list based on the evaluation performed by the calculation assistant. For example, assistant module 222 may determine, based on the weighted data, whether each recommendation included in the original recommendation list has been rejected by at least one of the calculation assistants. If a recommendation from the original recommendation list is not rejected by any of the calculation assistants based on the weighted data, the assistant module 222 may also include the recommendation in the improved reaped recommendation list Good.

  In response to the improved cropped recommendations list including at least one recommendation, the assistant module 222 may also output the modified cropped recommendations list for display to the user of the group Good. As one example, the assistant module 222 may cause the UI module 220 to display, via the display component 202, a GUI that includes the improved cropped recommendations list. As another example, the assistant module 222 may cause the UI module 220 to output, via the speaker component 208, the synthesized conversation including the improved clipped recommendations list. As another example, the assistant module 222 may output the improved cropped recommendation list to the computational assistant associated with the other users of the group, and the computational assistant then generates the improved cropped information. The suggested list may be displayed, spoken or output to its respective user.

  In some instances, the calculation assistants associated with the users of the group, including the assistant module 222, may use any of the data that is used to generate the recommendation list (clipped or improved and clipped) You may refrain from the output of instructions. For example, whether the calculation assistant is based on the second privacy level data (i.e. is the pruned list) or based on the weighted second privacy level data It may refrain from showing (i.e., whether it is the improved cropped list). The calculation assistant may refrain from outputting any indication of the original recommendation list. Thus, the calculation assistant may be able to generate a group recommendation based on the group member's private data without showing the private data. In this way, the calculation assistant may generate higher quality recommendations for the group.

  FIG. 3 is a block diagram illustrating an example computing system configured to execute an example virtual assistant in accordance with one or more aspects of the present disclosure. The assistant server system 360 of FIG. 3 will be described later as an example of the assistant server system 160 of FIG. FIG. 3 only illustrates one particular example of assistant server system 360, and many other examples of assistant server system 360 may be used in other instances and are included in exemplary assistant server system 360. It may include a subset of components or may include additional components not shown in FIG.

  As shown in the example of FIG. 3, assistant server system 360 includes one or more processors 340 of the user, one or more communication units 342, and one or more storage devices 348. Storage device 348 includes an assistant module 322, a search module 382, and user information data stores 325A through 325N (collectively, "user information data stores 325").

  Processor 340 is similar to processor 240 of computing system 210 of FIG. Communication unit 342 is similar to communication unit 242 of computing system 210 of FIG. Storage device 348 is similar to storage device 248 of computing system 210 of FIG. Communication channel 350 is similar to communication channel 250 of computing system 210 of FIG. 2, and thus may interconnect each of components 340, 342, and 348 for inter-component communication. In some instances, communication channel 350 may include a system bus, a network connection, an inter-process communication data structure, or any other method for communicating data.

  The search module 382 of the assistant server system 360 is similar to the search module 282 of the computing device 210 and may implement an integrated search function instead of the assistant server system 360. That is, search module 482 may perform a search operation instead of assistant module 322. In some instances, search module 382 may interface with an external search system to perform search operations instead of assistant module 322. When invoked, search module 382 may implement a search function, such as generating search queries and performing searches based on search queries generated across various local and remote information sources. The search module 382 may provide the results of the executed search to the launch component or module. That is, the search module 382 may output the search result to the assistant module 322.

  The assistant module 322 includes all functions of the local assistant module of the local assistant module 122 of FIG. 1 and the remote assistant module of the remote assistant module 123, and the assistant module 222 of the computing device 210 of FIG. May be included. Assistant module 322 may perform similar operations as the remote assistant module of remote assistant module 122 to provide an assistant service accessible via assistant server system 360. That is, assistant module 322 may act as an interface to a remote assistant service that can access computing devices in communication with assistant server system 360 over a network. For example, assistant module 322 may be an interface or API to the remote assistant module of remote assistant module 122 of assistant server system 160 of FIG.

  Similar to the user information data stores 124 and 125 of the system 100 of FIG. 1, the user information data stores 325 each include information that the assistant module 322 has learned or provided for each user. The personal records created by module 322 may be stored. The assistant module 322 may perform assistant tasks depending on the information stored in the data store 325 in addition to any information provided by the search module 382.

  FIG. 4 is a flow chart illustrating exemplary operations performed by one or more processors executing an exemplary virtual assistant, in accordance with one or more aspects of the present disclosure. FIG. 4 is described below in the context of the system 100 of FIG. For example, while local assistant module 122A is executing on one or more processors of computing device 110A, one or more of operations 402-420 in accordance with one or more aspects of the present disclosure may be performed You may implement. In some examples, while remote assistant module 122A is executing on one or more processors of assistant server system 160, one of operations 402-420 in accordance with one or more aspects of the present disclosure Or you may implement two or more. For the purposes of illustration only, FIG. 4 is described below in the context of computing device 110A of FIG.

  In operation, computing device 110A may receive a request to generate a recommendation for a group of users (402). As shown in the example of FIG. 1, the user A, the user B, and the user C may have a group chat. User A may ask "everyone still interested in taking a trip in March?", User B may respond by saying "I'm interested" and user C asks "where were you thinking?" It may be. However, looking at the flow of chat, user B may say "not this again, we can never agree on a place". To avoid having to discuss potential destinations, user A may say that they are "ask their virtual assistant for some recommendations". In the example of FIG. 1, user A may query the virtual assistant for some recommendations by entering an assistant activation command and provide a request for recommendations for the user's group. For example, the UI module 120 may receive an indication of user input at UID 112 of user A typing “@assistant where should we take a trip in March”. The UI module 120 may provide the user input to the assistant module 122A and / or 123A.

  An assistant module associated with the user of the group may obtain an agreement from the user of the group to use the first and second privacy level data (404). For example, assistant module 122A and / or 123A may obtain consent from user A to use the first and second privacy level data of user A, and assistant module 122B and / or 123B. May obtain consent from user B to use the first and second privacy level data of user B, and assistant modules 122C and / or 123C may receive first and second assistants of user C. Consent to use the privacy level data of may be obtained from user C.

  Assistant modules 122A and / or 123A may generate an original recommendation list for the group based on the first privacy level data for the users of the group (406). For example, assistant modules 122A and / or 123A may evaluate the first privacy level data for user A, user B, and user C to generate the original recommendation list. In this example, the first privacy level data for the group indicates that User A, User B, and User C all enjoy winter mountain sports (eg, ski or snowboarding), bourbon, and tropical vacation. It is also good. Based on the first privacy level data, assistant modules 122A and / or 123A are used in Bozeman, Montana for winter mountain sports, Lexington, Kentucky for Bourbon tasting and distillery tours, and Caribbean Islands for beach vacations. You may generate an original recommendation list that includes Assistant modules 122A and / or 123A may disseminate or output the original recommendation list to the calculation assistants associated with other users of the group. For example, assistant modules 122A and / or 123A may calculate the original recommendation list associated with user B (eg, assistant module 122B and / or 123B) and user C associated with the calculated assistant (eg, It may be output to the assistant module 122C and / or 123C).

  Each of the calculation assistants associated with the users of the group is based on the respective second privacy level data for the users of the group, and makes recommendations from the original recommendation list for inclusion in the clipped recommendations list. It may be evaluated (408). For example, assistant modules 122A and / or 123A determine whether user A has any negative preferences for recommendations from the original recommendation list based on the second privacy level data for user A You may Similarly, based on the second privacy level data for user B, the computational assistant associated with user B determines whether user B has any negative preference for recommendations from the original recommendation list. The calculation assistant associated with user C may determine that user C has some negative preference for recommendations from the original recommendation list based on the second privacy level data for user C. It may be determined whether or not.

  Continuing the above example, assistant modules 122A and / or 123A indicate that user A is in Montana based at least on the second privacy level data for user A indicating that user A has recently made a ski trip. It may be determined that it has a partially negative preference for winter mountain sports at Bozeman. Thus, the assistant modules 122A and / or 123A may determine that Bourbon tasting and distillery tours in Lexington, Kentucky and beach vacations in the Caribbean islands should be included in the cropped list, It may be determined that winter mountain sports in Bozeman, Montana should not be included in the cut-off list.

  The computational assistant associated with user B indicates that user B attempted to persuade a child and performed a heavy search for Zika virus Second privacy level data for user B and Zika virus to the Caribbean islands Based at least on the data indicating that it exists, it may be determined that user B has a completely negative preference for the Caribbean islands. Thus, the computational assistant associated with user B may determine that winter mountain sports in Bozeman, Montana and Bourbon tasting and distillery tours in Lexington, Kentucky should be included in the cut-off list, Caribbean It may be determined that beach vacations on the sea islands should not be included in the cut list.

  The computational assistant associated with user C indicates that user C is in Lexington, Kentucky, based at least on the second privacy level data for user C indicating that user C is not getting too much alcohol. It may be determined to have a partially negative preference for bourbon tasting and distillery tours. Thus, the computational assistant associated with user C may determine that winter mountain sports in Bozeman, Montana and beach vacations in the Caribbean islands should be included in the cut-off list, and in Lexington, Kentucky. It may be determined that bourbon tasting and distillery tours should not be included in the cut list.

  The assistant modules 122A and / or 123A may generate the cut recommendation list based on the evaluation performed by the calculation assistant. For example, assistant modules 122A and / or 123A may determine whether each recommendation included in the original recommendation list has been rejected by at least one of the calculation assistants. If a recommendation from the original recommendation list is not rejected by any of the calculation assistants, assistant module 122A and / or 123A may include the recommendation in the cut-off recommendation list.

  The assistant module 122A and / or 123A may determine 410 if there is a recommendation in the cut recommendations list for the result. In response to the clipped recommendations list including at least one recommendation (ie, at least one recommendation from the original recommendations list is not rejected by any of the calculation assistants associated with the users of the group If so ("yes" branch of 410), the assistant module 122A and / or 123A may output the cut recommendation list (412). As one example, the assistant modules 122A and / or 123A may cause the UI module 120 to display a GUI including the clipped recommendation list. As another example, the assistant modules 122A and / or 123A may cause the UI module 120 to output a synthesized conversation including the clipped recommendation list. As another example, assistant modules 122A and / or 123A may output the cut recommendations list to a calculation assistant associated with another user of the group, which calculation assistant is then The recommendation list may be displayed, spoken or output to its respective user.

  In response to the clipped recommendations list not including at least one recommendation ("No" branch of 410), each computational assistant associated with the users of the group is weighted for the users of the group Based on the respective second privacy level data, recommendations from the original recommendation list may be evaluated 414 for inclusion in the refined, clipped recommendation list. In some instances, the assessments may be performed in parallel. For example, each of the calculation assistants associated with the users of the group may, based on the respective second privacy level data for the users of the group, from the original recommendation list for inclusion in the cut recommendations list Recommendations from the original recommendation list to be evaluated in parallel and to be included in the improved reaped recommendation list, based on the weighted second privacy level data for the users of the group May be assessed. If the improved cropped list includes at least one recommendation ("yes" branch of 416), the assistant module 122A and / or 123A may output the improved cropped recommendation list ((4) 418). For example, assistant modules 122A and / or 123A may output the improved cropped recommendation list in a manner similar to the output of the cropped recommendation list described above. If the improved cropped list includes at least one recommendation ("No" branch of 416), then assistant module 122A and / or 123A may not have been able to provide a recommendation for the group by the calculation assistant An indication may be output (420).

  The original recommendation list continues the example above, including Bozeman, Montana for winter mountain sports, Lexington, Kentucky for Bourbon tasting and distillery tours, and the Caribbean islands for beach vacations. The assistant modules 122A and / or 123A may determine based on the evaluation that there are no recommendations remaining on the pruned list. In particular, winter mountain sports in Bozeman, Montana are denied by User A's computational assistant (ie, assistant modules 122A and / or 123A), and Bourbon Tasting and Distillery Tours in Lexington, Kentucky have User C's computational assistant ( That is, a beach vacation in the Caribbean islands was rejected by user B's calculation assistant (ie, assistant module 122B and / or 123B), which was rejected by the assistant modules 122C and / or 123C.

  The calculation assistants associated with user A, user B and user C are to be included in the improved cut-off recommendation list based on the respective weighted second privacy level data for that respective user You may evaluate the recommendations from the original recommendation list of For example, the second privacy level data for user A shows that user A has only a partially negative preference for winter mountain sports in Bozeman, Montana, Bourbon tasting and distillery tours in Lexington, Kentucky and Assistant Modules 122A and / or 123A are winter mountain sports in Bozeman, Montana, Bourbon tasting and distillery tours in Lexington, KY, as they show no negative preference for beach vacations in the Caribbean Islands. And a beach vacation on the Caribbean islands may be determined to be included in the improved cropped list. Similarly, the second privacy level data for User C shows that User C has partially negative preferences for Bourbon tasting and distillery tours in Lexington, Kentucky, and winter mountain sports and Bozeman, Montana. The computational assistant associated with user C is Winter Mountain Sports in Bozeman, Montana, Bourbon Tasting and Distillery Tour in Lexington, Kentucky, as it shows no negative preference for beach vacations in the Caribbean Islands. It may be determined that beach vacations in the Caribbean islands should be included in the improved cropped list. However, the second privacy level data for user B shows that user B has a completely negative preference for beach vacations in the Caribbean islands, winter mountain sports in Bozeman, Montana and Bourbon in Lexington, Kentucky • The computational assistant associated with user B is concerned with winter mountain sports in Bozeman, Montana and Bourbon tasting and distillery tours in Lexington, KY, as it shows no taste preference for tasting and distillery tours. It may be determined that it should be included in the improved cropped list, and it may be determined that beach vacations in the Caribbean islands should not be included in the improved cropped list.

  Assistant modules 122A and / or 123A should include Winter Mountain Sports in Bozeman, Montana and Bourbon tasting and distillery tours in Lexington, KY, in the improved cropped list based on the evaluation. It may be determined that Because the refined cropped list includes at least one recommendation, assistant module 122A and / or 123A may output the refined cropped list for display to users of the group.

  FIG. 5 is a flowchart illustrating exemplary operations performed by one or more processors executing an exemplary virtual assistant, in accordance with one or more aspects of the present disclosure. FIG. 5 is described below in the context of system 100 of FIG. For example, while local assistant module 122A is executing on one or more processors of computing device 110A, one or more of operations 502-524 in accordance with one or more aspects of the present disclosure may be performed You may implement. In some examples, while remote assistant module 122A is executing on one or more processors of assistant server system 160, one of operations 502-520 in accordance with one or more aspects of the present disclosure Or you may implement two or more. For purposes of illustration only, FIG. 5 is described below in the context of computing device 110A of FIG.

  In operation, computing device 110A may receive a request to generate a recommendation for a group of users (502). As shown in the example of FIG. 1, the user A, the user B, and the user C may have a group chat. User A may ask "everyone still interested in taking a trip in March?", User B may respond by saying "I'm interested" and user C asks "where were you thinking?" It may be. However, looking at the flow of chat, user B may say "not this again, we can never agree on a place". In order to avoid having to discuss potential destinations, user A may say that they ask "what's their ask for virtual assistant for some recommendations". In the example of FIG. 1, user A may query the virtual assistant for some recommendations by entering an assistant activation command, and provide a request for recommendations for the user to the group. For example, the UI module 120 may receive an indication of user input at UID 112 of user A typing “@assistant where should we take a trip in March”. UI module 120 may provide user input to assistant module 122A and / or 123A.

  The assistant module associated with the user of the group may obtain 504 consent from the user of the group to use the first and second privacy level data. For example, assistant module 122A and / or 123A may obtain consent from user A to use the first and second privacy level data of user A, and assistant module 122B and / or 123B. May obtain consent from user B to use user B's first and second privacy level data, and assistant module 122C and / or 123C may use user C's first and second Consent to use the privacy level data of 2 may be obtained from user C.

  The assistant modules 122A and / or 123A may generate the original recommendation list for the group based on the user of the first privacy level data for the group (506). For example, assistant modules 122A and / or 123A may evaluate the first privacy level data for user A, user B, and user C to generate the original recommendation list. In this example, the first privacy level data for the group indicates that User A, User B, and User C all enjoy winter mountain sports (eg, ski or snowboarding), bourbon, and tropical vacation. It is also good. Based on the first privacy level data, assistant modules 122A and / or 123A are used in Bozeman, Montana for winter mountain sports, Lexington, Kentucky for Bourbon tasting and distillery tours, and Caribbean Islands for beach vacations. You may generate an original recommendation list that includes Assistant modules 122A and / or 123A may disseminate the original recommendation list or output it to a calculation assistant associated with other users of the group. For example, assistant modules 122A and / or 123A may calculate the original recommendation list associated with user B (eg, assistant module 122B and / or 123B) and user C associated with the calculated assistant (eg, assistant -It may output to module 122C and / or 123C).

  The respective calculation assistants associated with the users of the group are based on the respective second privacy level data for the users of the group, from the original recommendation list to be included in the cut recommendation list May be assessed (508). For example, based on the second privacy level data for user A, assistant modules 122A and / or 123A determine whether user A has any negative preference for recommendations from the original recommendation list. You may Similarly, based on the second privacy level data for user B, the computational assistant associated with user B determines whether user B has any negative preference for recommendations from the original recommendation list. The calculation assistant associated with user C may determine that user C has some negative preference for recommendations from the original recommendation list based on the second privacy level data for user C. It may be determined whether or not.

  Continuing the above example, assistant modules 122A and / or 123A indicate that user A is in Montana based at least on the second privacy level data for user A indicating that user A has recently made a ski trip. It may be determined to have a partially negative preference (eg, 60% opposite preference) to winter mountain sports at Bozeman. Thus, the assistant modules 122A and / or 123A may determine that Bourbon tasting and distillery tours in Lexington, Kentucky and beach vacations in the Caribbean islands should be included in the cropped list, It may be determined that winter mountain sports in Bozeman, Montana should not be included in the cut-off list.

  Calculation Assistant Associated with User B Shows Second User Privacy Level Data for User B and the Zika Virus in the Caribbean Islands Indicates that User B Attempted to Persuade a Child and Perform a Mass Search About the Zika Virus Based on at least the data indicating that the user B is determined to have a completely negative preference (e.g., 100% opposite preference) to the Caribbean islands. Thus, the computational assistant associated with user B may determine that winter mountain sports in Bozeman, Montana and Bourbon tasting and distillery tours in Lexington, Kentucky should be included in the cut-off list, Caribbean It may be determined that beach vacations on the sea islands should not be included in the cut list.

  The computational assistant associated with user C indicates that user C is in Lexington, Kentucky, based at least on the second privacy level data for user C indicating that user C is not getting too much alcohol. It may be determined to have a partially negative preference (e.g., 25% opposite preference) for bourbon tasting and distillery tours. Thus, the computational assistant associated with user C may determine that winter mountain sports in Bozeman, Montana and beach vacations in the Caribbean islands should be included in the cut-off list, and in Lexington, Kentucky. It may be determined that bourbon tasting and distillery tours should not be included in the cut list.

  The assistant modules 122A and / or 123A may generate the cut recommendation list based on the evaluation performed by the calculation assistant. For example, assistant modules 122A and / or 123A may determine whether each recommendation included in the original recommendation list has been rejected by at least one of the calculation assistants. If a recommendation from the original recommendation list is not rejected by any of the calculation assistants, assistant module 122A and / or 123A may include the recommendation in the cut-off recommendation list.

  The assistant module 122A and / or 123A may determine 510 if there are any recommendations in the resulting cut recommendations list. In response to the cut recommendations list including at least one recommendation (ie, at least one recommendation from the original recommendations list was not rejected by any of the calculation assistants associated with the users of the group If) ("yes" branch of 510), the assistant module 122A and / or 123A may output the trimmed recommendations list (512). As one example, the assistant modules 122A and / or 123A may cause the UI module 120 to display a GUI including the clipped recommendation list. As another example, the assistant modules 122A and / or 123A may cause the UI module 120 to output a synthesized conversation including the clipped recommendation list. As another example, assistant modules 122A and / or 123A may output the cut recommendations list to a calculation assistant associated with other users of the group, and then the cut recommendations list It may be displayed, uttered or outputted to each user.

  In response to the clipped recommendations list not including at least one recommendation ("No" branch of 510), the respective calculation assistants associated with the users of the group are weighted for the users of the group Based on each of the second privacy level data, recommendations from the original recommendation list may be evaluated 514 for inclusion in the improved reaped recommendation list. For example, a computational assistant associated with a user of the group may determine whether the second privacy level data for the user indicates a negative preference for the recommendation above the initial threshold. In some instances, the threshold is initially set low (e.g., 5%, 10%, 15%, ..., 25%, etc.) and can not provide recommendations as described in more detail below. It may be increased as the case may be. The assistant may determine that the recommendations should not be included in the improved reaped recommendations list. In this case, the assistant determines that the second privacy level data for the user indicates that the user's preference for the recommendation is greater than the threshold. In some instances, if the assistant can not provide a recommendation using the initial threshold, the assistant may have the second privacy level data for the user be increased threshold (eg, 5 higher than the previous threshold). %, 10% etc.) may be determined to indicate that the user has a negative preference for the recommendation.

  If the improved cropped list includes at least one recommendation ("yes" branch of 516), the assistant module 122A and / or 123A may output the improved cropped recommendation list ((6) 518). For example, assistant modules 122A and / or 123A may output the improved cropped recommendation list in a manner similar to the output of the cropped recommendation list described above. If the improved cropped list includes at least one recommendation ("No" branch of 516), then assistant module 122A and / or 123A may use the weight at which the recommendation is evaluated to be above the maximum threshold. It may be determined if it is large (520). For example, assistant modules 122A and / or 123A determine whether the threshold used to evaluate the recommendation is greater than the maximum overall threshold level (eg, 75%, 85%, 90%, 95%) You may If the weight is greater than the maximum threshold ("yes" branch of 520), the assistant module 122A and / or 123A may output an indication that the calculation assistant could not provide any recommendation for the group ( 524). If the weight is not greater than the maximum threshold ("No" branch of 520), assistant module 122A and / or 123A may increase the weight (522). For example, assistant modules 122A and / or 123A may increase the threshold by 5%, 10%, 20%, etc. The respective computational assistants associated with the users of the group are based on the increased weighting of the respective second privacy level data for the users of the group in the improved cropped recommendation list. The recommendations from the original recommendations list for inclusion may be evaluated (514). For example, the computational assistant associated with the user of the group determines whether the second privacy level data for the user indicates that the user has a negative preference for the recommendation above the increased threshold. May be In some instances, the loop may continue until the calculation assistant can provide at least one recommendation. In some instances, the loop may continue until the threshold used to evaluate the recommendation is greater than the maximum overall threshold level. In this way, the assistant module may include the most desirable recommendations in the improved cropped list, without including recommendations for which some users have strong negative preferences thereto.

  The original recommendation list continues the example above, including Bozeman, Montana for winter mountain sports, Lexington, Kentucky for Bourbon tasting and distillery tours, and the Caribbean islands for beach vacations. The assistant modules 122A and / or 123A may determine based on the evaluation that there are no recommendations remaining on the pruned list. In particular, winter mountain sports in Bozeman, Montana are denied by user A's computational assistant (ie, assistant modules 122A and / or 123A), and Bourbon tasting and distillery tours in Lexington, Kentucky are user C (ie, assistant) The beach vacation in the Caribbean islands has been rejected by user B's calculation assistant (ie assistant module 122 B and / or 123 B), being rejected by the calculation assistant of modules 122 C and / or 123 C).

  The calculation assistants associated with user A, user B and user C are to be included in the improved cut-off recommendation list based on the respective weighted second privacy level data for that respective user You may evaluate the recommendations from the original recommendation list of For example, the calculation assistants associated with user A, user B, and user C have second privacy data for the respective users, wherein the respective users satisfy the initial threshold (10% for the description of this example) It may be determined whether it indicates that it has a preference for recommendations from the original recommendation list.

  For example, the second privacy level data for user A shows that user A has a 60% preference for winter mountain sports in Bozeman, Montana, and Bourbon tasting and distillery tours in Lexington, Kentucky and the Caribbean The assistant module 122A and / or 123A is a winter mountain sport in Bozeman, Montana (eg, 60% preference is greater than the 10% threshold, as it indicates no negative preference for beach vacations in the islands) ) Should not be included in the improved cropped list, and Bourbon tasting and distillery tours in Lexington, Kentucky and beach vacations in the Caribbean islands on the improved cropped list. Include It may be determined to be. Similarly, the second privacy level data for User C shows that User C has a 25% preference for Bourbon tasting and distillery tours in Lexington, Kentucky, and winter mountain sports and Caribbean in Bozeman, Montana. The computational assistant associated with user C has shown that he does not have a negative preference for beach vacations on the Sea Islands, and he is taking Bourbon tasting and distillery tours in Lexington, Kentucky (eg, the 25% preference is Determined that it should not be included in the improved cropped list, as it is greater than the 10% threshold, and winter mountain sports in Bozeman, Montana, and beach vacations in the Caribbean islands would be reaped. Should be included in the list It may be determined. In addition, the second privacy level data for user B shows that user B has 100% preference for beach vacations in the Caribbean islands, winter mountain sports in Bozeman, Montana and Bourbon in Lexington, Kentucky. The Calculation Assistant associated with User B improved the Winter Mountain Sports in Bozeman, Montana and the Bourbon Tasting and Distillery Tour in Lexington, Kentucky, as it shows no negative preference for tasting and distillery tours. It may be determined that it should be included in the harvested list and it may be determined that beach vacations in the Caribbean islands should not be included in the improved harvested list.

  The assistant modules 122A and / or 123A (for example, winter mountain sports in Bozeman, Montana are rejected by user A's calculation assistant and beach vacations in the Caribbean islands are user B's calculation assistant based on the evaluation) Denied that the recommendations from the original recommendations list should not be included in the improved reaped list, as it was rejected and because Bourbon tasting and distillery tours in Lexington, Kentucky were rejected by User C's calculation assistant) You may judge. The assistant module 122A and / or 123A may increase the threshold, as the improved cropped list does not include recommendations. For example, assistant module 122A and / or 123A may increase the threshold to 30%. As the 25% preference of User C for Bourbon tasting and distillery tours in Lexington, Kentucky is less than the increased 30% threshold, the computational assistant associated with User C is Bourbon tasting in Lexington, Kentucky And the distillery tour may be determined to be included in the improved reaped recommendations list. However, as the 60% preference of user A for winter mountain sports in Bozeman, Montana is greater than the increased 30% threshold, the computational assistant associated with user A continues to improve winter mountain sports in Bozeman, Montana. It may be determined that it should not be included in the cut-off recommendations list. Similarly, since User B's 100% preference for beach vacations in the Caribbean Islands is greater than the increased 30% threshold, the calculation assistant associated with User B is a beach vacation in the Caribbean Islands It may still be determined that it should not be included in the improved reaped recommendation list. Because the refined cropped list includes at least one recommendation, assistant module 122A and / or 123A may output the refined cropped list for display to users of the group.

  The following numbered examples may illustrate one or more aspects of the present disclosure.

  Example 1: In response to receiving a request for recommendation to a group of users, generating an original recommendation list for the group based on the first privacy level data for the users of the group The original recommendation list to be included in the cut recommendations list for the group, based on the respective calculation assistants associated with the users of the group and the respective second privacy level data for the users of the group Evaluating the recommendations from, the second privacy level being more restricted than the first privacy level, and responding to the step that the cut-off recommendations list includes at least one recommendation To display the clipped recommendation list for display to the users of the group. And a step of the method.

  Example 2: The cut-off recommendations list includes a first cut-off recommendations list, and the method responds to the first cut-off recommendations list not including a recommendation by the user of the group. Recommendations from the original recommendation list for inclusion in a second, clipped recommendations list for the group based on each computational assistant and on each of the weighted second privacy level data for the users of the group Outputting the second cropped recommendation list for display to a user of the group in response to the step of evaluating the second cropped recommendation list including at least one recommendation. The method according to Example 1, further comprising the steps of

  Example 3: Based on the calculation assistant associated with a particular user of the group and the respective second privacy level data for the particular user of the group, the first cropped recommendation list for the group The step of evaluating the recommendations from the original recommendation list for inclusion in each of the second privacy level data for a particular user of the group is such that the particular user has any negative preference for the recommendation. A method according to example 2, comprising determining not to include the recommendation in the first pruned recommendations list in response to determining that it indicates having the.

  Example 4: The second reaping for the group is performed by the calculation assistant associated with the particular user of the group and based on the respective weighted second privacy level data for the particular user of the group The step of evaluating the recommendation from the original recommendation list for inclusion in the selected recommendation list is that each second privacy level data for a particular user of the group is complete for the particular user to complete the recommendation. The method of example 3, comprising determining not to include the recommendation in the second pruned recommendation list in response to determining that it indicates a negative preference.

  Example 5: The second reaping for the group is performed by the computational assistant associated with the specific user of the group and based on the respective weighted second privacy level data for the specific user of the group The step of evaluating the recommendation from the original recommendation list for inclusion in the recommended list comprises: each second privacy level data for a particular user of the group, the particular user being a partial for the recommendation The method of any combination of Examples 1 to 4 comprising determining that the recommendation is to be included in the second reaped recommendation list in response to determining that it indicates a negative preference for .

  Example 6: The request is received from a requesting user not included in the group of users, and the original recommendation list is not generated based on data for the requesting user, as described in any combination of examples 1 to 5. Method.

  Example 7: The method according to any combination of examples 1 to 5, wherein the request is received from a requesting user included in the group of users, and the original recommendation list is generated based on data for the requesting user .

  Example 8: The method of any combination of Examples 1 to 7, wherein the request is for one or more of entertainment recommendations, food recommendations, travel recommendations, and shared product recommendations.

  Example 9: The method of any combination of examples 1 to 8, wherein the first privacy level data for the users of the group comprises data publicly available to the users of the group.

  Example 10: At least one processor and, when executed, the at least one processor responsive to receiving a request for a recommendation to a group of users and a first privacy level for the users of the group Based on the data, generate an original recommendation list for the group, output the original recommendation list to each calculation assistant associated with the user of the group, and calculate each of the calculations associated with the user of the group Receive an indication from the assistant, based on the respective second privacy level data for the users of the group, whether or not the recommendations from the original recommendation list should be included in the cut-off recommendations list for the group In response to the clipped recommendations list including at least one recommendation, A computing device comprising: at least one memory comprising instructions for providing a computational assistant configured to output a list of recommendations that has been removed for display to a user of the group; A computing device wherein the privacy level of 2 is more restrictive than the first privacy level.

  Example 11: The cut-off recommendations list includes the first cut-off recommendations list, and the calculation assistant is configured to calculate each weighted assistant for the users of the group from the calculation assistants associated with the users of the group. Receiving an indication, based on the second privacy level data, whether the recommendations from the original recommendation list should be included in the second clipped recommendations list for the group, and the second clipped recommendations The computing according to Example 10, further configured to output the second cropped recommendation list in response to the list including at least one recommendation and for display to a user of the group. ·device.

  Example 12: A computational assistant associated with a particular user may have each second privacy level data for a particular user in the group that the particular user has some negative preference for the recommendation The computing device according to Example 11, configured to determine not to include recommendations from the original recommendation list in the first pruned recommendations list in response to determining to indicate.

  Example 13: The computational assistant associated with the particular user indicates that each second privacy level data for the particular user of the group indicates that the particular user has a complete negative preference for the recommendation. The computing device of example 12, configured to determine not to include the recommendation in the second pruned recommendation list in response to determining to indicate.

  Example 14: A computational assistant associated with the particular user indicates that the respective second privacy level data for the particular user of the group has the particular user a partially negative preference for the recommendation. The computing device according to any combination of examples 10-13, configured to determine to include the recommendation in the second pruned recommendation list in response to determining that it is indicated.

  Example 15: The request is received from a requesting user who is not included in the group of users, and the original recommendation list is not generated based on data for the requesting user, as described in any combination of examples 10 to 14. Computing device.

  Example 16: The computer according to any of the combinations of examples 10 to 14, wherein the request is received from a requesting user included in the group of users, and the original recommendation list is generated based on data for the requesting user. Device.

  Example 17: The computing device according to any combination of Examples 10 to 16, wherein the request is for one or more of entertainment recommendations, food recommendations, travel recommendations, and shared product recommendations. .

  Example 18: The computing device according to any combination of examples 10-17, wherein the first privacy level data for the users of the group comprises data publicly available to the users of the group.

  Example 19: When executed, one or more processors of a computing device, in response to receiving a request for a recommendation to a group of users, and a first privacy level for users of that group Based on the data, generate an original recommendation list for the group, output the original recommendation list to each calculation assistant associated with the user of the group, and calculate each of the calculations associated with the user of the group Receive an indication from the assistant, based on the respective second privacy level data for the users of the group, whether or not the recommendations from the original recommendation list should be included in the cut-off recommendations list for the group In response to the clipped recommendations list including at least one recommendation, Non-transitory computer readable storage medium storing instructions for outputting the recommended list to the user of the group, the second privacy level being more restricted than the first privacy level Non-transitory computer readable storage medium.

  Example 20: The cut-off recommendations list includes a first cut-off recommendations list, the non-transitory computer readable storage medium is to the one or more processors of the computing device, to the users of the group From the associated respective calculation assistant, based on the respective weighted second privacy level data for the users of the group, recommendations from the original recommendations list are taken as the second clipped recommendations list for the group Received an indication of whether or not to be included in the second pruned recommendations list in response to the second pruned recommendations list including at least one recommendation and for displaying to a user of the group Non-transitory computer readable according to Example 19, further storing instructions for outputting a recommendation list憶媒 body.

  Example 21: The non-transitory computer readable storage medium according to example 19, further comprising instructions that cause the one or more processors to perform the method described in any combination of examples 1-9.

  Example 22: A computing device comprising means for performing the method described in any combination of examples 1-9.

  In one or more examples, the functions described may be implemented in hardware, software, firmware, or any combination thereof. When implemented in software, the functions may be stored on or transmitted over as one or more instructions or code on a computer readable medium and executed by a hardware based processing unit. Computer readable media may include computer readable storage media or media. The medium corresponds to a tangible medium such as a data storage medium, or a communication medium including any medium that facilitates transfer of a computer program from one place to another, eg according to a communication protocol. Thus, a computer readable medium may generally correspond to (1) a tangible computer readable storage medium that is non-transitory, or (2) a communication medium such as a signal or carrier wave. A data carrier may be any available to one or more computers or one or more processors to retrieve instructions, code and / or data structures for implementing the techniques described in this disclosure. Media may be used. The computer program product may include computer readable media.

  By way of example and not limitation, such computer readable storage media may be in the form of RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage device, flash memory, or instructions or data structures. It can include any other storage medium that can be used to store the desired program code and can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, instructions may use wireless technologies such as coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or infrared, wireless, and microwave from a website, server, or other remote source When transmitted, such coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, wireless, and microwave are included in the definition of medium. However, it should be understood that computer readable storage media and media and data storage media relate to non-transparent, tangible storage media without connection, carrier waves, signals or other transparent media. Disks, as used herein, include compact disks (CDs), laser disks, optical disks, digital versatile disks (DVDs), floppy disks and Blu-ray® disks. The disk normally reproduces the data magnetically and the laser optically reproduces the data. Combinations of the above should be included within the scope of computer readable media.

  The instructions may be one or more digital signal processors (DSPs), general purpose microprocessors, application specific integrated circuits (ASICs), field programmable logic arrays (FPGAs), or other equivalent integrated circuits or discrete logic circuits May be executed by one or more processors such as Thus, the term "processor" as used herein may refer to any of the foregoing structures or any other structure suitable for implementing the techniques described herein. Further, in some aspects, the functionality described herein may be provided in dedicated hardware and / or software modules. Also, the techniques can be fully implemented with one or more circuits or logic elements.

  The techniques of the present disclosure may be implemented in a wide variety of devices or apparatuses, including a wireless handset, an integrated circuit (IC) or a set of ICs (eg, a chip set). Although various components, modules, or units are described in this disclosure as highlighting functional aspects of devices configured to implement the disclosed technology, implementations with different hardware units are not necessarily required. It is not necessary. Rather, as mentioned above, the various units couple to hardware units or interoperate hardware units comprising one or more processors as described above in connection with appropriate software and / or firmware. May be provided by a set of

  Various embodiments have been described. These and other embodiments are within the scope of the following claims.

110A-110N Computing Device 120 UI Module 122A Local Assistant Module 123A-123N Remote Assistant Module 124A User Information Data Store 125A-125N User Information Data Store 130 Network 160 Assistant Server System 180 Search Server System 182 Search Module 202 Display Component 204 Presence Detection Input Component 206 Microphone Component 208 Speaker Component 210 Computing Device 212 User Interface Device 220 UI Module 222 Assistant Module 224 User Information Data Store 226 Application Module (s)
240 processor (s)
242 Communication unit (s)
244 Input Component (s)
246 Output Component (s)
248 storage device 250 communication channel 282 search module 322 assistant module 325A-325N user information data store 340 processor (s)
342 Communication Unit (s)
348 storage device 350 communication channel 360 assistant server system 382 search module

Claims (20)

A method implemented by one or more processors executing a computational assistant associated with each user belonging to the group,
A step of in response to receiving the recommended request to the group of the user, based on the first privacy level data for the users in the group, generates the original recommendation list to the group,
The original to be included in the cut-off recommendation list for the group by respective computational assistants associated with the users of the group and based on respective second privacy level data for the users of the group Evaluating recommendations from the recommendations list, wherein the second privacy level is more restricted than the first privacy level;
Outputting the clipped recommendations list for display to the users of the group in response to the cropped recommendations list including at least one recommendation;
Method, including. The clipped recommendations list includes a first clipped recommendations list, and the method
In response to the first pruned recommendations list not including a recommendation, a weighted second privacy level for each user of the user of the group and for each user of the group. Evaluating recommendations from the original recommendation list for inclusion in a second reaped recommendation list for the group based on the data;
Outputting the second cropped recommendations list for display to the users of the group in response to the second cropped recommendations list including at least one recommendation;
The method of claim 1, further comprising To be included in the first cropped recommendation list for the group by the computational assistant associated with the particular user of the group and based on respective second privacy level data for the particular user of the group Evaluating the recommendations from the original recommendation list of
In response to determining that each second privacy level data for a particular user of the group indicates that the particular user has some negative preference for the recommendation, the recommendation is 3. The method of claim 2, including the step of: determining not to include in the first cropped recommendations list. The second cropped recommendations for the group by the computational assistant associated with the particular user of the group and based on respective second privacy level data weighted for the particular user of the group Evaluating the recommendations from the original recommendation list for inclusion in the list
Said recommendation in response to determining that each second privacy level data for a particular user of said group indicates that said particular user has a complete negative preference for said recommendation. 4. The method of claim 3, including the step of: determining that it is not included in the two trimmed recommendations list. The second cropped recommendations for the group by the computational assistant associated with the particular user of the group and based on respective second privacy level data weighted for the particular user of the group Evaluating the recommendations from the original recommendation list for inclusion in the list
Said recommendation in response to determining that each second privacy level data for a particular user of said group indicates that said particular user has a partially negative preference for said recommendation. 4. The method of claim 3, including the steps of: determining to include in the second reaped recommendations list.   The method according to claim 1, wherein the request is received from a requesting user not included in the group of users, and the original recommendation list is not generated based on data for the requesting user.   The method according to claim 1, wherein the request is received from a requesting user included in the group of users, and the original recommendation list is generated based on data for the requesting user. The request is
Recommended entertainment,
Food recommendations,
The method of claim 1, wherein the method is for one or more of travel recommendations and shared product recommendations.   The method of claim 1, wherein the first privacy level data for the users of the group comprises data publicly available to the users of the group. At least one processor,
When executed, said at least one processor,
Based on the first privacy level data for the user in response and the group that received the recommended requirements of the users of the group, to produce the original recommendation list to the group,
In the calculation assistant respectively associated with the user of the group, to output the source of the recommended list,
The recommendations from the original recommendation list are clipped to the group from the respective calculation assistants associated with the users of the group based on respective second privacy level data for the users of the group or to receive an indication whether or not that should be included in the recommended list,
Instructions for providing a computational assistant configured to output the clipped recommendations list for display to the user of the group in response to the cropped recommendations list including at least one recommendation A computing device comprising: at least one memory;
The second privacy level is more limited than the first privacy level.
Computing device. The cut-off recommendations list includes a first cut-off recommendations list, and the calculation assistant
Recommendations from the original recommendation list for the group based on respective weighted second privacy level data for the users of the group from respective computational assistants associated with the users of the group. Receive instructions on whether to include in the second reaped recommendations list,
Further configured to output the second cropped recommendations list in response to the second cropped recommendations list including at least one recommendation and for display to the user of the group. 11. The computing device of claim 10. The calculation assistant associated with a particular user is
The original in response to determining that each second privacy level data for a particular user of the group indicates that the particular user has some negative preference for the recommendation. It constituted the recommendations from the recommendation list for determines not included in the first of clippings recommendation list, computing device of claim 11. The calculation assistant associated with the particular user is
Said recommendation in response to determining that each second privacy level data for a particular user of said group indicates that said particular user has a complete negative preference for said recommendation. The computing device according to claim 12, configured to determine not to be included in the two trimmed recommendations list. The calculation assistant associated with the particular user is
Said recommendation in response to determining that each second privacy level data for a particular user of said group indicates that said particular user has a partially negative preference for said recommendation. Determine to include in the second reaped recommendations list,
The computing device of claim 12 configured to:   11. The computing device of claim 10, wherein the request is received from a requesting user not included in the group of users, and the original recommendation list is not generated based on data for the requesting user.   11. The computing device of claim 10, wherein the request is received from a requesting user included in the group of users, and the original recommendation list is generated based on data for the requesting user. The request is
Recommended entertainment,
Food recommendations,
11. The computing device of claim 10, for one or more of travel recommendations and shared product recommendations.   11. The computing device of claim 10, wherein the first privacy level data for the users of the group comprises data publicly available to the users of the group. Non-transitory computer readable storage medium having instructions recorded thereon, comprising:
The instructions when executed on one or more processors of a computing device
Generating an original recommendation list for the group in response to receiving a request for a recommendation for the group of users and based on first privacy level data for users of the group,
Outputting the original recommendation list to each calculation assistant associated with the users of the group;
The recommendations from the original recommendation list are clipped to the group from the respective calculation assistants associated with the users of the group based on respective second privacy level data for the users of the group an indication of whether that should be included in the recommended list is received,
Outputting the clipped recommendations list for display to the user of the group in response to the cropped recommendations list including at least one recommendation ;
The second privacy level is more limited than the first privacy level.
Non-transitory computer readable storage medium. The cropped recommendations list includes a first cropped recommendations list, and the non-transitory computer readable storage medium is to the one or more processors of the computing device.
Recommendations from the original recommendation list for the group based on respective weighted second privacy level data for the users of the group from respective computational assistants associated with the users of the group. Receive instructions on whether to include in the second reaped recommendations list,
Further recording instructions causing the second cropped recommendations list to be output in response to the second cropped recommendations list including at least one recommendation and for display to the user of the group 20. The non-transitory computer readable storage medium of claim 19.

Images